Fatimah Wilson is part of a social experiment under way in Richmond, Calif., an economically depressed corner of the San Francisco Bay Area.

Wilson is pregnant and is spending the day with other soon-to-be moms learning habits— from better eating to relaxation—to help them improve their health and the health of their babies. The goal: to erase the health disparity that results in African American infants in Contra Costa County dying at twice the rate of white babies before they reach their first birthday.

Wilson, 34, attended the West County African American Community Baby Shower, where she ate healthful food, received gifts for her baby and mingled with other local women. The women attended workshops where they learned, among other things, to use yoga-based relaxation techniques to reduce stress during pregnancy.

“We can change the statistics, because they are gross and grave throughout the United States for African American women and babies,” says Lynor Jackson- Marks, one of the organizers of the shower. Nationally, African American babies are more than twice as likely as white babies to die before their first birthday.

Reducing disparities and educating mothers-to-be were the event’s laudable goals. But years of research suggest that teaching mothers-to-be such as Wilson tips for a healthy pregnancy will only go so far. The gap in the rates of infant mortality can’t be explained by unhealthful behaviors. Instead, it is part of a pattern that goes back generations and persists despite an individual’s changes in income, environment, behavior and living conditions. It is a puzzle with no easy solution—and one that is almost certainly beyond what typical prevention efforts can achieve.

The Rise of Prevention

Preventive programs became popular over the past few decades, as public health officials focused on the difference in health status among racial and ethnic groups, economic classes and geographic locations. That focus reflected researchers’ new understanding of the close connection between health and social factors such as income and race, and is part of a larger movement to address preventable deaths.

“The idea that these are health disparities really emerged about 30 years ago,” explains Nancy Adler, professor of psychiatry at the UCSF School of Medicine and chair of the MacArthur Research Network on Socioeconomic Status and Health. The fact that they are avoidable and hit poor people and people of color harder is what distinguishes a disparity in health from a difference in health. “These differences,” Adler explains, “are avoidable and unjust.”

About 40 percent of deaths in the United States are attributable to avoidable illnesses such as heart disease, analysis by the Institute of Medicine has shown. Among the avoidable deaths, as Adler notes, are all of those caused by disparities.

Public health departments, traditionally focused on preventing communicable diseases, have begun shifting more of their resources to prevention. Contra Costa County, where Richmond is located, has targeted disparities for a decade. The federal Affordable Care Act also reflects the sea change in moving toward improving the health of people by reducing preventable disease, earmarking a huge sum—$10 billion—for prevention initiatives. Healthy People 2020, on ongoing federal initiative to reduce chronic illness and preventable death, names reducing disparities in health as one of their primary goals. They intend to achieve that goal by “Integrating prevention into the continuum of education—from the earliest ages on,” according to the program’s guiding framework.

Changes in behavior, researchers have come to think, can save lives, an understanding that has reshaped public health policy.

The puzzle of infant mortality

The focus on individual behavior as a way to address health disparities seems to suggest that differences in health behavior cause disparities. In the case of infant mortality, for instance, events such as the community shower might imply that African American mothers have bad habits that white mothers do not share. But that’s not actually true, researchers have found. Pregnant African American women, for instance, do not smoke more than other pregnant women or engage in other behaviors linked to higher infant mortality rates in numbers sufficient to explain the higher death rates of their babies.

The problem is much harder to untangle: It is the result of a lifelong diminishment of health that starts before birth and is passed on through generations. On close examination, what has been treated as a problem of an individual’s life choices emerges instead as a deeply rooted social problem.

“What we are seeing in differences between blacks and whites is not just a result of what is happening during the nine months of pregnancy, but actually has also to do with what happens prior to pregnancy,” says Dr. Neal Halfon. Halfon is a professor in the departments of pediatrics, health sciences and policy studies at UCLA, the director of the Center for Healthier Children, Families and Communities, and a former policy advisor to former Vice President Al Gore.

His groundbreaking 2003 article on disparities in infant mortality (co-authored with UCLA colleague Michael Lu) featured an illustration imagining the effects of circumstances on health as a series of upward and downward pressures over the course of a lifetime. Positive events boost health, and stressful, negative events hurt health. Because of poverty and discrimination, African American women often experience more stressful life events and fewer positive ones compared to white women, and as a result their health suffers. So does the health of their children, starting with their development in the womb.

During different periods of development, we are more or less sensitive to our environment. A baby’s initial development in the womb is a critical time. Other important moments occur in childhood. Trauma and stress during these periods affect health permanently. Chronic stress outside these critical periods of development hurts health too. The cumulative effect of stress and disadvantage over the mother’s lifetime, Halfon explains, affects the health of their child.

Some of the stress African American women feel comes from living in disadvantaged neighborhoods where violence and uncertainty in housing and employment are facts of life. And African Americans are poor at sharply disproportional rates. More than 27 percent of African Americans were poor in 2010, compared to about 10 percent of whites.

Health is closely related to income. As income levels for African Americans rose between 1968 and 1978 following the civil rights movement, for instance, mortality rates for African Americans declined. When African American income started to fall again in comparison to white income in the 1980s, the gap between the mortality rates of the two groups grew once again.

The relationship between race, income and health is seen specifically in infant mortality too. Researchers at the National Bureau of Economic Research, for instance, found a relationship between income and low birth weight. Raising the incomes of single, high school educated mothers by as little as $1,000 reduces rates of low birth weight, a predictor of infant mortality, by about 7 to 11 percent. The biggest improvements, they found, are among African American mothers.

Researchers have been trying to unearth the causes of these relationships for years. They do know that the feeling of being a part of an isolated group—one that other social groups view as distinct and below them on the social hierarchy—is a part of the experience of poverty that’s harmful to health. That connection was revealed by a ground- breaking study of British civil servants in the late 1960s, called the Whitehall Study, which showed a social gradient in health. People at the top of the social hierarchy have the best health, and people at the bottom have the worst. People in the middle of the hierarchy, who do not lack access to care and have sufficient incomes, also have worse health than those at the top of the hierarchy. The social gradient affects everyone’s health—and that may be the primary culprit in the poor health of low-income people.

Amani Nuru-Jeter, a professor of public health at the University of California, Berkeley, says that people understand when their place on the social ladder is on the lowest rung. “People know when they are living in those kinds of neighborhoods,” she explains. “And knowing that can be stressful.”

But the puzzle is even more complicated than that. African American babies who are not born to poor mothers are also more likely to die within their first of life than white babies, suggesting that the effects of poverty linger past the day when a woman is no longer poor. That fact also suggests that race affects health whether or not you are poor. African American mothers with a college education—an indicator of higher socioeconomic status—have infant mortality rates of 10 per 100,000 births. That’s three times higher than rates for babies born to white mothers with a college education.

For African American mothers, stress is “ever present in your life, because of how you are treated as a member of a racial minority in this country,” Halfon says. Kids are often aware of their status as a minority from a young age. “Children who experience the kind of racism that has to do with their status in society feel that status,” he says. “That can just wear against them over long periods of time.”

A telling fact that supports the relationship between discrimination and health is the lower death rates of babies born to African immigrants. African immigrants who are new to the United States have similar birth outcomes to those of white women. The children of African immigrants, however, have birth outcomes similar to those of African American women—a pattern not seen in white immigrants. The data suggest that something particular to living as a black woman in the United States is hurting the health of their children.

The experience of poverty and the stress of occupying a lower rung of the social ladder may be the cause of the disproportionate share of health problems borne by African Americans. But thinking about health in that way—as a social problem rather than an individual problem, and a systemic problem rather than a health behavior problem—requires a conceptual shift in thinking. And it means that a solution to health disparities will require much more than the current emphasis on prevention programs.

Global Approaches

People live longer on average in Sweden and Norway than they do in the United States, as do people in 49 other countries. The United States ranks 50th in life expectancy from birth, a number that is attributable at least in part to sharp health disparities, including the infant mortality rate.

Halfon points to the example of the protective social programs in Scandinavian countries, where education is equalized and unemployment benefits are generous enough to keep citizens from feeling that they will fall into an abyss if they lose their jobs. Social programs don’t kick in only after an individual is in acute distress, as they do here.

“We pay when people fail,” Halfon says. “Other countries invest for success—and invest for equity.” They have figured out that early investments produce social dividends for everyone. Our approach, he adds, lacks that kind of consideration. “If NASA used the same kind of philosophy that we used in social programs,” he says, “they would launch satellites into any old trajectory and spend all of their money to make sure they didn’t fall to the ground.” In light of the scale of disparities in problems like infant mortality, solutions that rely on individuals making different choices are unlikely to work.

Laurette Dubé, the founding chair and scientific director of the McGill World Platform for Health and Economic Convergence, is one expert pushing for a change of thinking about global health. Dubé’s work on the interconnection between systems such as the agriculture industry and worldwide problems such as hunger and obesity appeared in a recent special issue of Proceedings of the National Academy of Science.

Behavior is linked to health, as current approaches to reducing preventable illness suggest. But so are social and economic systems. “Right now public health experts are saying we should behave differently than we do,” Dubé says, “and they are right.” But, she adds, “if the whole machine is going 300 miles per hour in a direction that runs counter to the change we need to be making, we will never make any significant dent in the changes that need to be made.”

Improving health requires changes to entire systems, Dubé says. Public health plays an integral role in improving population health, but prevention efforts must be more closely tied to widespread re form to truly improve health. “It is clear that we need a whole social change.”

Change, Dubé stresses, is “critical.” She points to escalating health-care costs and their ever-increasing share of national budgets. Health-care expenditures in the United States, for instance, doubled between 2000 and 2010, according to analysis by the Centers for Medicare and Medicaid Services. “In industrialized countries,” Dubé says, “we are reaching the limits of what financially society can afford in terms of health care.”

Despite the urgency, sweeping changes do not seem to be on the horizon. “I’ve been seeing more that’s distressing,” Adler says. “It seems like we are going in the wrong direction on this issue.”

“We may not have the political will to do it,” she adds. “But disparities are preventable over time if we would make that a priority. We could drastically reduce them.”

Nuru-Jeter of UC Berkeley agrees that larger policy changes would help reduce health disparities. But since that is unlikely, smaller interventions remain important. “I definitely think we need broader scale society reformation,” she says. “In the meantime, we don’t just sit still and wait for that to happen.”

At the baby shower in Richmond, no one was simply waiting for better health to happen. Instead, the women carefully followed directions from an instructor to breathe in and out at a late-morning stress management class, doing what they could to improve their own health and the health of their children.

Prevention programs have been successful in the past. For example, they have been shown to sharply reduce smoking and the illnesses associated with tobacco, and disparities in infectious diseases like the flu were reduced decades ago. In short, prevention has made huge improvements in the health of the U.S. population. These programs, however, have dealt with problems that are simpler to solve, with a clearer cause and effect. But disparities such as those that affect infant health and mortality today are far more complex. They are a profound public health problem, but they may also be a problem that public health can’t solve.